admirelurk

admirelurk t1_ja9wy95 wrote

I counter that many developers of ML have a too narrow definition of what constitutes danger. Sure, chatGPT will not go rogue and start killing people, but the technology affects society in much more subtle ways that are hard to predict.

6